Homocysteine concentration and adenosine A2A receptor production by peripheral blood mononuclear cells in coronary artery disease patients

Abstract Hyperhomocysteinemia is associated with coronary artery disease (CAD). The mechanistic aspects of this relationship are unclear. In CAD patients, homocysteine (HCy) concentration correlates with plasma level of adenosine that controls the coronary circulation via the activation of adenosine A2A receptors (A2AR). We addressed in CAD patients the relationship between HCy and A2AR production, and in cellulo the effect of HCy on A2AR function. 46 patients with CAD and 20 control healthy subjects were included. We evaluated A2AR production by peripheral blood mononuclear cells using Western blotting. We studied in cellulo (CEM human T cells) the effect of HCy on A2A R production as well as on basal and stimulated cAMP production following A2A R activation by an agonist‐like monoclonal antibody. HCy concentration was higher in CAD patients vs controls (median, range: 16.6 [7‐45] vs 8 [5‐12] µM, P < 0.001). A2A R production was lower in patients vs controls (1.1[0.62‐1.6] vs 1.53[0.7‐1.9] arbitrary units, P < 0.001). We observed a negative correlation between HCy concentration and A2A R production (r = −0.43; P < 0.0001), with decreased A2A R production above 25 µM HCy. In cellulo, HCy inhibited A2AR production, as well as basal and stimulated cAMP production. In conclusion, HCy is negatively associated with A2A R production in CAD patients, as well as with A2A R and cAMP production in cellulo. The decrease in A2A R production and function, which is known to hamper coronary blood flow and promote inflammation, may support CAD pathogenesis.


| INTRODUC TI ON
Homocysteine (HCy) is a thiol-containing amino acid intermediate the metabolism of which is linked to those of methionine, uric acid and adenosine. HCy and adenosine concentrations are correlated at least in coronary artery disease (CAD) patients. 1 Adenosine is an ATP derivative that is released by endothelial cells and myocytes during ischaemia, hypoxia or inflammation. 2,3 Adenosine impacts the cardiovascular system via the activation of its receptors namely A 1 R, A 2A R, A 2B R or A 3 R pending on their pharmacological properties. 4 A 2A R activation exerts artery vasodilation via cAMP production 5 and specially coronary vasodilation, 6 cAMP production and coronary vasodilation being correlated. 7 Chronic ischaemia elicits coronary vasodilation in the myocardium. 8 This adaptive response is partly due to adenosine release that improves coronary blood flow via activation of A 2A R 6,9,10 and A 2B R. 9,10 Although acute release of adenosine leads to coronary vasodilation and may be consequently beneficial for the myocardium, chronic exposure to high adenosine level may have deleterious effects. 11 Hyperhomocysteinemia (HHCy) is associated with cardiovascular disease 12,13 and independently associated with CAD 14,15 and myocardial infarction-induced death. 16 It was also found that HHCy is correlated with CAD severity 1,17 although the mechanistic aspects of this relationship are unclear. Finally, HHCy is associated with increased oxidative stress 18 and endothelial dysfunction. 19 However the precise mechanism by which HHCy participates into CAD progression remains controversial because homocysteine-lowering therapy does not affect the inflammatory status of CAD patients 20 and poorly influences cardiovascular risk. 21 Peripheral blood mononuclear cells (PBMC) can be easily sampled and the behaviour of adenosine receptors produced on PBMC mirrors their counterparts in heart, 22 and in coronary arteries. 23 Therefore, PBMCs are useful to address A 2A R pharmacological properties in patients and to evaluate the influence of HCy on adenosine receptors of the vascular system. We previously observed a decrease in A 2A R production by PBMC of CAD patients 24 the cause of which not yet investigated.
Thus, we evaluated here the relationship between HCy concentration and A 2A R production in CAD patients. We also examined in cellulo the effect of HCy concentration on A 2A R and cAMP production in basal conditions and after agonist exposure.

| Panel size
A difference in APL or A 2A R expression >20%-25% was considered to have pathophysiological relevance; accordingly, a panel of 30-40 subjects was considered to be sufficient to provide statistical signification.

| Study population
We recruited 46 patients (13 women and 33 men, mean age 69.3 ± 11.6 years) admitted for coronary angiography in the department of Cardiology, University Hospital, Marseille, between January 2016 and January 2018. Clinical presentation could be an acute coronary syndrome or stable angina. The patients included presented with a significant CAD defined by an angiographic stenosis ≥50%. Exclusion criteria for the study were creatinine clearance <25 mL/min and age <18 or >80 years. Twenty healthy subjects (7 women and 13 men, mean age 63 ± 7) from the medical staff, matched for age and sex, were used as controls for the adenosinergic profile. They were without history of cardiovascular or inflammatory disease and not under any

| Adenosine concentration measurement
Adenosine plasma concentration (APC) as well as adenosine measurement in free cell culture supernatant were performed as previously described by liquid chromatography-tandem mass spectrometry after extraction 26,27 using a Shimadzu UFLC XR system (Shimadzu, Marne la Vallée, France).
Plasma extraction procedure: internal standard solution with 2-Chloro adenosine was prepared (300 nM in water). Plasma sample (100 µL) was transferred into a microfuge tube. Each sample was spiked with internal standard solution (50 µL) and methanol (300 µL) and vortexed for 1 minute. Samples were then centrifuged (4°C, 10 minutes, 13 300 g). Supernatant was then evaporated to dryness at 60°C under nitrogen. Formic acid (0.1% in water; 150 µL) was then added and quickly vortexed prior to transfer in an HPLC auto-sampler vial. The intra-assay coefficient of variation (CV) was <10%.

| HCy measurement
Blood was collected in a tube with EDTA and centrifuged (4°C, 10 minutes, 2000 g). Plasma was frozen and stored until assay.
Total homocysteine was quantified with the LC-MS Clinmass® 'Homocysteine in plasma/serum' kit (Recipe, Germany).
Supernatants were analysed using a Shimadzu UFLC XR system consisting of two LC-20ADXR binary pumps, a DGU20A5R vacuum degasser, a CT0-20AC thermostated column oven and a SIL-20ACXR cooled auto sampler (Shimadzu). The LC system was interfaced with an ABSciex 4500 triple quadrupole mass spectrometer (Les Ulis, France) operating with an electrospray ionization source (ESI) using nitrogen (purity: 99.99%). The intra-or inter-assay CV was <5%.

| PBMC A 2A R production
The procedure has been described. [28][29][30] In brief, blood was col- (AU) as previously described 31,32 (the ratio of pixels generated by the A 2A R band to pixels generated by the background signal was calculated). In these conditions, the intra-or inter-assay CV was <10%.

| Adenosine concentration measurement in cell culture medium
Cell-free supernatant was blotted on a Whatmann blot paper (6 mm diam.) prior to extraction using a mixture of methanol (400 µL) and internal standard (see above) for 90 minutes at 45°C.
After extraction, an aliquot (350 μL) was evaporated to dryness at 60°C under nitrogen. Formic acid (0.1% in water; 150 µL) was added and vortexed prior to transfer into an HPLC auto-sampler vial. Dosage was then performed using LC-MS/MS as described above.

| Adenosine deaminase activity (ADA) measurement
Ado (28 mM; 750 μL) was mixed with cell culture medium (750 μL) in NaCl 0.9% (2 mL final volume). Aliquots were then incubated (40 minutes, 37°C). The reaction was started by adding the substrate and was stopped by cold immersion. COBAS 8000 apparatus (Roche®, Geneva, Switzerland) was used to quantify ammonia concentration. The intra-and inter-assay coefficients of variation ranged between 3% and 5%.

| cAMP dosage
The method has been previously described. 25

| Statistical analysis
Data were described by mean and standard deviation or median and interquartile range. Correlations between biological parameters were quantified and tested using Pearson's correlation coefficient.

Comparisons of biological parameters between patients and controls
were performed using a variance analysis (ANOVA two ways). All statistical tests were two-sided and P values less than 0.05 were considered statistically significant. Analysis were performed using the SPSS software (version 13.0 2004; SPSS Inc, Chicago, IL, USA).
In patients, we observed a negative correlation between HCy concentration and A 2A R production (r = −0.43; P < 0.0001; Figure 3A), with decreased A 2A R production above 25 µM HCy ( Figure 3B). We did not find a correlation between C reactive protein and A 2A R production (r = 0.16, P = 0.41) while a trend in correlation was found between HCy and C reactive protein concentrations (r = 0.33, P = 0.06).

| D ISCUSS I ON
We report here a negative correlation between HCy concentration and A 2A R production by PBMC in CAD patients. We also found that HCy concentrations in the range of those measured in CAD patients decreased both A 2A R and cAMP production in CEM human T cells.
In the latter conditions, it is likely that down regulation of A 2A R resulted from HCy and not from changes in adenosine concentration or ADA.
These results support and expand our previous observation that took advantage of an in cellulo model of inflammation and hypoxia, which may be considered as reflecting CAD conditions. Using CEM T cells, we previously showed that HCy decreases A 2A R and cAMP production via H 2 S and NF kappa B pathway. 33,34 In patients, low A 2A R and cAMP production may participate in CAD in three ways: (a) by altering the adaptive vasodilation of coronary arteries when oxygen supply is needed, an hypothesis that is supported by the correlation found between the CAD gravity score (Syntax score) and HCy concentration 1 ; (b) by inhibiting the adenosinergic T cell immunosuppression mechanism via H 2 S production that, in turn, promotes inflammation 33 ; and (c) by favouring platelet aggregation and activation through HCy/H 2 S pathway, thus contributing to atherothrombosis, stroke or myocardial infarction. 35 It was reported that an increase in 5 µM HCy promotes the incidence of CAD 36 and that HHCy is associated with restenosis of CAD patients treated by percutaneous intervention 37 as well as with cardiovascular causes of death. 38 Furthermore, an association between CRP and HCy concentrations was reported in patients with acute myocardial infarction. 39 The effects of HCy in CAD may result from an increase in oxidative stress in the vascular endothelium. 40 It was shown that HCy promotes endothelial cell dysfunction via the up-regulation of p66shc expression following hypomethylation of the promoter. 41 51,52 Conversely, the decrease in A 2A R activation found in CAD patients probably promotes inflammation that in turn promotes atherosclerosis. 53,54 In summary, we found that HCy concentrations measured in CAD patients are associated with low production of A 2A R in CAD patients as well as with low production of cAMP in cellulo. These data are consistent with the possibility that HCy participates in CAD pathophysiology by reducing coronary blood flow and by promoting inflammation and atherosclerosis.

| Limitations
The possible influence of A 1 and/or A 2 B receptors on the effects of HCy on cAMP production was not addressed in this study.

| CON CLUS ION
HCy is negatively correlated with A 2A R production in CAD patients, and negatively associated with A 2A R production and cAMP level in cellulo. The decrease in A 2A R production and cAMP level, which is known to hamper coronary blood flow and promote inflammation, may support CAD pathogenesis.